Society & Work
How many jobs are taken by automation every day?
The roughly 600,000 industrial robots installed in 2025 account for 960,000 to 3.6 million displaced worker-equivalents, and nobody can say more precisely, because jobs rarely vanish in one stroke. They disappear through attrition: someone leaves and the position is never refilled. The estimate rests on 2020 research that found each robot per 1,000 workers lowers employment by 0.2 percentage points in affected regions. Zoom out and the forecast is 92 million jobs displaced globally by 2030, about 18 million a year, or one every couple of seconds on paper.
Roughly 35 jobs every minute.
Source: WEF Future of Jobs 2025; Acemoglu & Restrepo (MIT) 2020; IFR 2025.
Last checked: 18 October 2026
Which jobs are genuinely at risk from automation - and which are safer than you think
The Oxford Martin School's landmark 2013 study predicted that 47% of US jobs were at high risk of automation within 20 years. That prediction sparked enormous debate and has been partially revised. The actual displacement rate has been slower than predicted, partly because automating complex, judgment-dependent tasks is harder than anticipated, and partly because humans adapt by shifting to new roles. But the direction of travel is clear and measurable.
The jobs most vulnerable are those with high routine content, whether manual or cognitive. Assembly line work, data entry, basic bookkeeping, and scripted customer service are all highly susceptible. Genuinely safer are roles requiring physical dexterity in unpredictable environments (plumbers, electricians), complex interpersonal judgment (nurses, therapists), and creative problem-solving with ambiguous parameters (engineers, architects). The pattern is not about education level: it is about routine versus non-routine task content.
Acemoglu and Restrepo's research also shows that technological change historically creates new jobs, just not always for the same workers or in the same places. The near-term challenge is geographic and demographic concentration: manufacturing workers in their 50s in industrial regions face a much harder adjustment than younger workers in service sectors. An 18 million per year aggregate masks very different experiences depending on where you are and what you do.
Automation & job displacement by the numbers
WEF Future of Jobs 2025: 92 million jobs displaced by 2030 (~18.4M/year); robotics/autonomous systems is the largest net job displacer
MIT Acemoglu & Restrepo (2020): each robot per 1,000 workers reduces employment by 0.2 percentage points
Industrial robots alone displace an estimated 960,000-3.6M worker-equivalents per year at current installation rates
McKinsey: 15% of all work tasks globally can be automated with current technology (2024)
Jobs taken by robots vs. robots deployed, today
Each new robot deployed pushes the displacement count higher. The gap between those two numbers is the question economists have been arguing over for a decade.
From early mechanization to AI: the automation timeline
- 2013Oxford Martin 47% automation risk paper triggers global policy debate
- 2020MIT causal confirmation: robots do displace workers (Acemoglu & Restrepo)
- 2025WEF Future of Jobs 2025: 92M displaced by 2030, 170M created; robotics/autonomous systems is largest net displacer
Robot-driven job displacement over time
MIT research confirmed in 2020 that each robot per 1,000 workers reduces employment by a measurable 0.2 percentage points. WEF Future of Jobs editions have long published gross displacement projections in the same order of magnitude: 75M (2018), 85M (2020), 83M (2023), then 92M by 2030 in the 2025 report. The live counter uses that 2025 baseline, about 18.4 million a year, or 35 jobs a minute.
| Year | Rate | Jobs/day | Context |
|---|---|---|---|
| 2018 | 36/min | 52K | Manufacturing automation; pre-AI wave |
| 2020 | 32/min | 46K | COVID drives automation acceleration |
| 2023 | 39/min | 56K | First FoJ net-negative creation/destruction balance |
| 2025 | 35/min | 50K | Structural macrotrends; robotics/autonomous systems largest net displacer |
Robots vs. workers: the evidence on industrial automation and jobs
Measuring automation displacement
The cleanest empirical evidence is Acemoglu and Restrepo's 2020 Journal of Political Economy study, which used IFR robot installation data matched to US commuting zone employment records. Regions that received more industrial robots showed lower employment and wages, and the effect was causal, not just correlated. The effect size is moderate per unit: 0.2 percentage points per robot per 1,000 workers. Applying a range of 1.6 to 6 displaced jobs per robot (the paper's own aggregate estimate is about 3.3) to 540,000 new industrial robots installed in 2024 gives roughly 860,000 to 3.2 million worker-equivalents per year (540,000 × 1.6 ≈ 864,000; 540,000 × 6 ≈ 3.24 million). Software automation and AI add more that is harder to measure.
Industrial robots vs. software automation
The public debate focuses on factory robots, but software-based automation likely displaces more jobs. RPA tools like UiPath and Automation Anywhere automate finance, HR, and back-office tasks with no hardware. AI-powered systems replace call centre agents without any robot in sight. Displacement increasingly happens in office settings rather than on factory floors, which makes it harder to trace and politically harder to address: there is no empty factory floor to photograph.
Economic research: Acemoglu, WEF, and the debate on automation
| Year | Finding | Value | Source |
|---|---|---|---|
| 2013 | Oxford Martin (Frey & Osborne): 47% of US occupations at high risk of automation within 20 years | 47 % US occupations at high automation risk | World Economic Forum |
| 2020 | Acemoglu & Restrepo (MIT): each robot per 1,000 workers causes 0.2% employment reduction; confirms causal displacement | 0.2 % employment reduction per robot per 1,000 workers | World Economic Forum |
| 2025 | WEF Future of Jobs 2025: 92M jobs displaced by 2030 (~18.4M/year); 170M created; net +78M. Robotics/autonomous systems is the largest net job displacer | 92.0M jobs displaced projected by 2030 | World Economic Forum |
In perspective
At the WEF average, 35 jobs disappear from the projection every minute. One every couple of seconds, on a factory floor, in a warehouse, or at a customer service desk.
Eighteen million displacements a year is nearly twice the employed population of the Netherlands.
The WEF's 92 million by 2030 is larger than Germany's entire labour force.
Frequently asked questions
- How many jobs are taken by robots per year?
- WEF Future of Jobs 2025 projects 92 million jobs displaced by 2030 from structural macrotrends, about 18.4 million a year. Academic estimates using IFR installation data put industrial robot displacement alone at 960,000 to 3.6 million direct manufacturing jobs per year, depending on methodology. The WEF total is broader than factory robots: it covers technology-driven structural change across roles, including software and AI.
- Does each robot really displace multiple workers?
- Acemoglu and Restrepo (2020) put it at 0.2 percentage points of employment reduction per robot per 1,000 workers. Other studies estimate 1.6 to 6 job displacements per robot installed, with the range driven by sector and task-overlap. Industrial robots in repetitive assembly work displace more jobs per unit than robots in variable-task environments.
Cite this rate
AnythingCounter derived rate, last checked October 2026: about 35 jobs per minute (about 50,400 jobs per day).
Method: Based on World Economic Forum 2025, converted to a per-minute rate (daily total ÷ 1,440).
Suggested citation: AnythingCounter, "Jobs taken by robots," October 2026, https://anythingcounter.com/automation-statistics
Primary source: World Economic Forum.
How this number is calculated, and how much to trust it
The live counter uses WEF Future of Jobs 2025: 92 million jobs displaced by 2030 from structural macrotrends. Averaged over five years that is 18.4 million/year ÷ 525,600 minutes ≈ 35/min, or one job every ~1.7 seconds. The Acemoglu & Restrepo (MIT, 2020) figure of 0.2 pp employment reduction per robot per 1,000 workers is the academic cross-check for industrial robots specifically, not a separate global total.
Academic grounding for robot-specific causality is the 2020 Journal of Political Economy paper by Acemoglu and Restrepo, "Robots and Jobs: Evidence from US Labor Markets," which matched IFR installation data to US commuting-zone employment. FoJ displacement totals swing with each survey round; treat the annualised rate as a projection average, not a measured global headcount.
Sources: see below. More detail on our methodology page.
Sources
Worth a look... Robots deployed - Jobs lost to automation - Automation statistics.